Flip a Coin
Flip one to ten virtual coins for heads or tails — a fair 50/50 from your browser’s secure random generator, with a running tally, streaks and the exact odds.
Flip a Coin
How to use the coin flipper
Choose how many coins to toss (1 to 10) and press Flip, tap a coin, or press the space bar. Gold is heads and silver is tails, and the result is also written out in words and read aloud by screen readers. The tally counts every coin you toss, tracks your current and longest streak, and shows the last 60 results in order. Reset tally starts again. The page address keeps your last flip and your best streak, so Share this result sends a friend a link that shows your flip and challenges them to beat your streak. If your device is set to reduce motion, the coins change without spinning.
How the flip is decided
For each coin the page asks crypto.getRandomValues — the cryptographically secure random generator built into
your browser — for a whole number from 0 to 4,294,967,295. Even numbers mean heads, odd numbers mean tails. Because exactly
half of the 4,294,967,296 possible values are even, the chance of heads is exactly one half, with no rounding and no bias,
and each coin is independent of every other coin and every earlier flip. With several coins, the number of heads follows the
binomial distribution:
P(heads) = 2,147,483,648 ÷ 4,294,967,296 = 1/2
P(exactly k heads with n coins) = C(n, k) ÷ 2ⁿ
C(n, k) = n! ÷ (k! × (n − k)!) Worked example: three coins
Three coins can land in 2 × 2 × 2 = 8 equally likely orders: HHH, HHT, HTH, THH, HTT, THT, TTH and TTT. Exactly two heads appears in three of them (HHT, HTH, THH), so C(3, 2) = 3 and the chance is 3 ÷ 8 = 37.5% — the same number the working shows when you flip 3 coins and get 2 heads. All heads or all tails is 1 ÷ 8 = 12.5% each. With 10 coins, exactly 5 heads — the “expected” result — happens only 252 ÷ 1,024 = 24.61% of the time: most flips of ten coins are not an even split, and the odds table under the tally shows every possibility.
Streaks are normal
Runs of the same side feel suspicious, but they are exactly what randomness looks like. The tally shows the chance of a streak at least as long as your best one in the number of tosses you have made, computed exactly with a recursion of the kind described by Mark Schilling in The Longest Run of Heads (1990). Counting streaks of either side:
| Tosses | Average longest streak | Chance of a streak of at least… |
|---|---|---|
| 10 | 3.66 | 4 in a row: 46.48% |
| 20 | 4.66 | 5 in a row: 45.84% |
| 50 | 5.98 | 6 in a row: 54.41% |
| 100 | 6.98 | 7 in a row: 54.23% |
| 200 | 7.98 | 8 in a row: 54.19% |
Each doubling of the number of tosses adds about one to the typical longest streak. A streak does not change what happens next: after any run, the next toss is still 50/50.
Are real coin tosses fair?
Almost, but not quite. In 2007 the mathematicians Persi Diaconis, Susan Holmes and Richard Montgomery analyzed the physics of a hand-caught toss and predicted that a coin lands on the same side it started about 51% of the time, because a flipped coin wobbles as it spins. A team led by František Bartoš tested this with 350,757 flips by 48 people, published in the Journal of the American Statistical Association in 2025. Coins landed on their starting side 50.8% of the time (95% credible interval 50.6% to 50.9%), while heads overall came up 50.0% — so heads and tails were balanced, but the starting side had a small edge that varied between flippers. A virtual flip has no starting side and no wobble, so it is exactly 50/50. For another quick random decision, try the dice roller.
Frequently asked questions
Is this online coin flip really 50/50?
Yes. Each coin is decided by a random whole number from your browser’s secure generator, crypto.getRandomValues: even means heads, odd means tails. Exactly half of the 4,294,967,296 possible numbers are even, so each side has a chance of exactly 1 in 2, and no flip depends on the ones before it.
Is heads or tails more likely with a real coin?
Neither, on average. In a 2025 study of 350,757 hand-caught flips by 48 people, heads came up 50.0% of the time. The same study found a small same-side bias: a coin landed on the side that was facing up at the start 50.8% of the time. A virtual coin has no starting side, so that bias does not apply here.
What are the odds of flipping heads 10 times in a row?
1 in 1,024 (about 0.098%) if you name heads before the first flip. But long streaks of either side are more common than most people expect: in 100 flips there is an 8.67% chance of a streak of 10 or more somewhere, and a streak of at least 7 is more likely than not (54.23%).
After five heads in a row, is tails “due”?
No — this is the gambler’s fallacy. The coin has no memory, so the next flip is still exactly 50/50. Six heads in a row is unlikely (1 in 64) only when judged before any flips; once five heads have happened, the sixth has the usual 1 in 2 chance.
How do I use a coin flip to settle a decision fairly?
Agree who gets heads and who gets tails before pressing Flip, and do not re-flip a result you dislike. For more than two options use the spin the wheel picker, and for numbers the random number generator.
Sources
- Bartoš et al. (2025) — Fair coins tend to land on the same side they started: evidence from 350,757 flips, Journal of the American Statistical Association
- Diaconis, Holmes & Montgomery (2007) — Dynamical bias in the coin toss, SIAM Review 49(2)
- Schilling (1990) — The longest run of heads, The College Mathematics Journal 21(3)
- W3C — Web Cryptography Level 2 (getRandomValues)